Anorexia nervosa (AN) is a chronic and often fatal disorder that affects 0.3% of women. There is no FDA-approved treatment, and the mortality rate is 5% per decade. In addition to environmental influence, family and twin studies demonstrate substantial heritability for AN. Because the etiology of this devastating illness is not known, we undertook a pilot study to examine its genetic underpinnings. With the support of a private foundation, our multicenter collaboration has collected 196 multiplex AN kindreds from 7 sites across North America and Europe. With a limited sample, this pilot study has produced four suggestive linkages from a genome-wide scan, one very close to genome-wide significance (Chromosome 1 at 70 cM, p - 0.0001; Chr. 1 at 202 cM, LOD = 3 46. p = 0.00003; Chr. 2 at 102 cM, LOD = 2.22: p = 0.00070; and Chr. 13 at 102 cM. LOD = 2.50; p = 0.00035) The first suggestive linkage results from a subset of the sample, namely individuals with the restricting subtype of AN. The other results were obtained by incorporating two covariates, drive-for-thinness from the Eating Disorders lnventory-2 and the total score from the Yale-Brown Obsessive Compulsive Scale, into covariate-based linkage analysis. Based on these very promising linkage findings, we believe genes underlying liability to AN can be mapped by augmenting the pilot sample. Thus support is requested for a multicenter effort to collect 400 affected relative pairs with AN. Over a five year period, the 11 collaborating groups (10 clinical, 1 analytic) will collect diagnostic and other phenotypic data and blood samples from 400 multiplex AN kindreds. The UNIVERSITY OF PITTSBURGH is one of these research groups. each of which is submitting a nearly identical application as a group of collaborating R01s. Microsatellites will be genotyped at H 10 cM intervals across the genome using all new families. Linkage analyses will be conducted by using diagnostic and phenotypic data to confirm suggestive linkages from the pilot study and to identify new genomic regions of interest. The diagnostic and genetic data and lymphoblastoid cell lines (derived from blood samples) will become part of a national archival resource for genetic studies of AN through the NIMH Genetics Initiative.